Preventing recurrence of deterministic failures

1. A method, comprising:

receiving, by a management console managing a server, a request to perform an operation on the server, wherein the operation comprises a set of steps;

determining that performance of the set of steps of the operation is associated with a deterministic failure in the server;

determining that the deterministic failure is of a first class of deterministic failures, of a plurality of classes of deterministic failures; and

blocking performance of the set of steps of the operation by the management console based on the first class of failure, thereby preventing an occurrence of the deterministic failure in the server.

2. The method of claim 1 , wherein the operation is a system management operation, further comprising:

prior to receiving the request:

receiving an earlier request to perform the set of steps of the system management operation on the server;

performing the set of steps of the system management operation on the server;

determining that performing at least one step of the set of steps of the system management operation on the server causes the deterministic failure in the server;

determining that the deterministic failure is of the first class of deterministic failures; and

storing, by the management console, an indication that: (i) the system management operation is a prohibited operation, and (ii) the deterministic failure is of the first class of deterministic failures.

3. The method of claim 2 , wherein the first class of the plurality of classes of deterministic failures comprises a cosmetic failure that does not generate incorrect results, wherein a second class of the plurality of classes of deterministic failures comprises a permanent failure that occurs for all input values provided to one of the set of steps of the operation, wherein a third class of the plurality of classes of deterministic failures comprises a recoverable failure that the server recovers from without operator intervention, wherein a fourth class of the plurality of classes of deterministic failures comprises an unrecoverable failure that requires restarting the server to correct the failure, wherein a fifth class of the plurality of classes of deterministic failures comprises a transient failure that occurs based on a set of input values provided to one of the set of steps of the operation, the method further comprising:

sharing, by the management console managing the server, the indication with a second management console, wherein the second management console manages a second server.

4. The method of claim 3 , further comprising:

outputting, by the management console, an indication that performance of the operation has been blocked because the operation is associated with the first class of deterministic failure;

outputting an indication that the first class of deterministic failure is the cosmetic failure; and

responsive to receiving input specifying to perform the set of steps of the operation, performing the set of steps of the operation.

5. The method of claim 1 , wherein determining that the operation is associated with the deterministic failure comprises:

identifying, in a data store storing a set of prohibited operations, an entry associated with the operation.

6. The method of claim 5 , further comprising:

upon determining that performance of the operation no longer results in the deterministic failure, removing the entry associated with the operation from the data store.

7. The method of claim 1 , wherein the operation is a system management operation, wherein the server is part of a computing cluster comprising a plurality of servers, wherein a plurality of management consoles manage a respective subset of the plurality of servers in the cluster, wherein the request is issued by at least one of: (i) a user, and (ii) a management application configured to provide high availability of the server in the cluster.

8. A system, comprising:

a server; and

a management console configured to manage the server, wherein the management console is configured to perform an operation comprising:

receiving, by a management console managing a server, a request to perform an operation on the server, wherein the operation comprises a set of steps;

determining that performance of the set of steps of the operation is associated with a deterministic failure in the server;

determining that the deterministic failure is of a first class of deterministic failures, of a plurality of classes of deterministic failures; and

blocking performance of the set of steps of the operation by the management console based on the first class of failure, thereby preventing an occurrence of the deterministic failure in the server.

9. The system of claim 8 , wherein the operation is a system management operation, wherein the operation further comprises:

prior to receiving the request:

receiving an earlier request to perform the set of steps of the system management operation on the server;

performing the set of steps of the system management operation on the server;

determining that performing at least one step of the set of steps of the system management operation on the server causes the deterministic failure in the server;

determining that the deterministic failure is of the first class of deterministic failures; and

storing, by the management console, an indication that: (i) the system management operation is a prohibited operation, and (ii) the deterministic failure is of the first class of deterministic failures.

10. The system of claim 9 , wherein the first class of the plurality of classes of deterministic failures comprises a cosmetic failure that does not generate incorrect results, wherein a second class of the plurality of classes of deterministic failures comprises a permanent failure that occurs for all input values provided to one of the set of steps of the operation, wherein a third class of the plurality of classes of deterministic failures comprises a recoverable failure that the server recovers from without operator intervention, wherein a fourth class of the plurality of classes of deterministic failures comprises an unrecoverable failure that requires restarting the server to correct the failure, wherein a fifth class of the plurality of classes of deterministic failures comprises a transient failure that occurs based on a set of input values provided to one of the set of steps of the operation, the operation further comprising:

sharing, by the management console managing the server, the indication with a second management console, wherein the second management console manages a second server.

11. The system of claim 10 , the operation further comprising:

outputting, by the management console, an indication that performance of the operation has been blocked because the operation is associated with the first class of deterministic failure;

outputting an indication that the first class of deterministic failure is the cosmetic failure; and

responsive to receiving input specifying to perform the set of steps of the operation, performing the set of steps of the operation.

12. The system of claim 8 , wherein determining that the operation is associated with the deterministic failure comprises:

identifying, in a data store storing a set of prohibited operations, an entry associated with the operation.

13. The system of claim 12 , the operation further comprising:

upon determining that performance of the system management operation no longer results in the deterministic failure, removing the entry associated with the operation from the data store.

14. The system of claim 8 , wherein the operation is a system management operation, wherein the server is part of a computing cluster comprising a plurality of servers, wherein a plurality of management consoles manage a respective subset of the plurality of servers in the cluster, wherein the request is issued by at least one of: (i) a user, and (ii) a management application configured to provide high availability of the server in the cluster.

15. A computer program product, comprising:

a non-transitory computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code comprising:

computer-readable program code configured to receive, by a management console managing a server, a request to perform an operation on the server, wherein the operation comprises a set of steps;

computer-readable program code configured to determine that performance of the set of steps of the operation is associated with a deterministic failure in the server;

computer-readable program code configured to determine that the deterministic failure is of a first class of deterministic failures, of a plurality of classes of deterministic failures; and

computer-readable program code configured to block performance of the set of steps of the operation by the management console based on the first class of failure, thereby preventing an occurrence of the deterministic failure in the server.

16. The computer program product of claim 15 , wherein the operation is a system management operation, further comprising:

computer-readable program code configured to, prior to receiving the request:

receive an earlier request to perform the set of steps of the system management operation on the server;

perform the set of steps of the system management operation on the server;

determine that performing at least one step of the set of steps of the system management operation on the server causes the deterministic failure in the server;

determine that the deterministic failure is of the first class of deterministic failures; and

store, by the management console, an indication that: (i) the system management operation is a prohibited operation, and (ii) the deterministic failure is of the first class of deterministic failures.

17. The computer program product of claim 16 , wherein the first class of the plurality of classes of deterministic failures comprises a cosmetic failure that does not generate incorrect results, wherein a second class of the plurality of classes of deterministic failures comprises a permanent failure that occurs for all input values provided to one of the set of steps of the operation, wherein a third class of the plurality of classes of deterministic failures comprises a recoverable failure that the server recovers from without operator intervention, wherein a fourth class of the plurality of classes of deterministic failures comprises an unrecoverable failure that requires restarting the server to correct the failure, wherein a fifth class of the plurality of classes of deterministic failures comprises a transient failure that occurs based on a set of input values provided to one of the set of steps of the operation, the computer program product further comprising:

computer-readable program code configured to share, by the management console managing the server, the indication with a second management console, wherein the second management console manages a second server.

18. The computer program product of claim 17 , further comprising:

computer-readable program code configured to output, by the management console, an indication that performance of the operation has been blocked because the operation is associated with the first class of deterministic failure;

computer-readable program code configured to output an indication that the first class of deterministic failure is the cosmetic failure; and

computer-readable program code configured to, responsive to receiving input specifying to perform the set of steps of the operation, perform the set of steps of the operation.

19. The computer program product of claim 15 , wherein determining that the system management operation is associated with the deterministic failure comprises:

identifying, in a data store storing a set of prohibited operations, an entry associated with the operation.

20. The computer program product of claim 19 , further comprising:

computer-readable program code configured to, upon determining that performance of the system management operation no longer results in the deterministic failure, remove the entry associated with the operation from the data store.